Dear authors,

there are 2 anonymous reveiws which you should carefully address in all points. In particular, I share the view that it is not a total column paper but your results are only relevant for the stratosphere and this should be addressed via a changed title.

Furthermore, please carefully address these additional major concerns from my side:

- I do not agree with your arguments in reply to Box 1.5 The linear fit in Figure 6 does not mean much, other than as a baseline, as there are two linear regimes, one from 07:00 to 13:00 and from 13:00 to 16:00 hours. Is there an explanation for the two regimes?
- Fig. 3a in Sussmann et al. 2005 does not show indications for any measurable non-linear diurnal change
- Your modeled diuarnal variation is much closer to a linear behavior than your measurement results. This means, obviously the main reason for your observed non-linear behavior is a different (probably experimental) one, i.e., your trying to explain via modeling is flawed or can at best be used as a minor, partial explanation. So you should throughly discuss possible reasons which could make the measurements erroneously look non-linear, i.e., please quantitatively estimate possible measurement articfacts like airmass/zenith angle dependencies.

Thanks for successfully performing all suggested improvements. Congratulations on an interesting study.
Sincerely,
Ralf Sussmann